An alternative approach to vector vibrating sample magnetometer

REVIEW OF SCIENTIFIC INSTRUMENTS VOLUME 69, NUMBER 9 SEPTEMBER 1998

An alternative approach to vector vibrating sample magnetometerdetection coil setup

E. O. Samwela)

Digital Measurement Systems Division of ADE Technologies, Inc., Newton, Massachusetts 02166

T. Bolhuis and J. C. LodderMesa Research Institute, University of Twente, 7500 AE Enschede, the Netherlands

~Received 5 February 1996; accepted for publication 22 May 1998!

Vector vibrating sample magnetometers~VSMs! can present problems with respect to angulardependent calibration and positional dependency when they are used for measurements on thin filmsamples, which have dimensions comparable to or larger than the sample–coil distances. Theproblems are due to the fact that in conventional VSMs the sample is rotating with respect to thecoils, when performing angular dependent measurements. In this article a solution is presentedbased on a setup of VSM detection coils, whose position is linked to that of the sample. Togetherwith a newly designed sample holder, the above mentioned problems are prevented or reduced. Thevector detection coil system shows a relatively small error in the determination of the magnetizationvector ~61% in the absolute value and60.6° in the angle!. Furthermore, it has a relatively smallpositional dependency~1% per mm! combined with a sufficient sensitivity~1 nA m2 or 1 memu at10 s time constant! and a capability of using samples up to 10310 mm2. The improved sampleholder for thin film measurements reduces positional problems while, at the same time, reducing thebackground signals of the holder~to 10 pA m2 per kA/m or 7.958310210 emu/Oe!. © 1998American Institute of Physics.@S0034-6748~98!04008-8#

th

esystngltohu

iointi

o-

inhenga

roeuli

ropytrin-

ntseter

n

ldonersub-een

theiew

re-illrgethehenhascedne-lm

, antematedcan

I. INTRODUCTION

The medium motion along the recording head duringrecording process causes a continuous change in bothlength and the direction of the applied field vector with rspect to the medium. Therefore, a vector measurementtem should be used for the analysis of recording mediathat the magnetization vector is measured rather thanprojection of the magnetization on the field direction. In geeral, for all media where the field is applied under an anwith the anisotropy direction, the magnetization vecshould be measured. This presents more information tscalar measurements. An example of a vectorial measment ~on a standardg-Fe2O3 audio tape at 80°! is given inFig. 1. This figure shows the change in length and directof the magnetization vector as a function of the changapplied field. From the maximum field, where the magnezation aligns with the field, the magnetization will rotate twards the easy axis direction~which is in plane! for decreas-ing fields. Close to the field where theM vector lengthreaches its minimum, the particles start flipping, changtheir magnetization irreversibly. From that point onward, tmagnetization vector will again align with the increasinegative field. This type of figure presents a lot of informtion on the actual reversal process that is not available fscalar measurements. To obtain this type of result, a msurement system with a sensor capable of vectorial measments is required. An overview of several very useful app

a!Electronic mail: [email protected]

3200034-6748/98/69(9)/3204/6/$15.00

Downloaded 04 Oct 2001 to 130.89.34.231. Redistribution subject to A

ethe-s-o

he-eranre-

ng-

g

-ma-re--

cations of vector measurement systems, such as anisotmeasurements, recording simulation experiments, and insic hysteresis loops, is given in Ref. 1.

One very commonly used instrument for measuremeon magnetic materials is the vibrating sample magnetom~VSM!, which was introduced in 1956 by Foner,2 who wrotea more detailed paper in 1959.3 Other authors such as vaOosterhout,4 Plotkin,5 and Flanders6 used a similar instru-ment but vibrated the sample in the direction of the fierather than perpendicularly as has been described by Fand has since become the standard. In Ref. 3 and manysequent publications various detection coil setups have bproposed for uniaxial and biaxial signal detection. One ofbest setups is presented in Ref. 1, together with an overvof several other vectorial detection coil systems.

Traditionally VSMs have been used mostly for measuments on fairly small samples. However, most VSMs wshow calibration problems if samples that are relatively lain one or more dimensions compared to their distance todetection coils are used. This problem becomes visible wthe sample is rotated or when the magnetization vectoran angle with the applied field. With the growing importanof thin film recording media for magnetic hard disk anvideotape applications, the need arises for vectorial magtometers capable of measuring relatively large thin fisamples with low magnetic moments. Bernards1 and Richter7

have published systems suiting these needs. In this articlealternative approach is presented. The detection coil syspresented here does not show the need for a compliccalibration procedure as described in Refs. 7 and 8 andbe used with samples up to 131 cm2 while it is at the same

4 © 1998 American Institute of Physics

IP license or copyright, see http://ojps.aip.org/rsio/rsicr.jsp

sedroa

ityksseon

oh

pe

thela

r

it-

eled

eas

eno

poeca

itee

wlifise

ri-y

msto

uxngestheec-the

eld

llyas

uen-nsator

yst

tionnd ad-tlylowthe

ryas-o are-le,rs,

orms

htsecednti-

tes a

tew

ctiion

3205Rev. Sci. Instrum., Vol. 69, No. 9, September 1998 Samwel, Bolhuis, and Lodder

time somewhat more sensitive than both of these systemIn addition, an improved sample holder will be describ

for thin film samples, which improves measurement repducibility and drastically decreases the background signcaused by the sample holder and the sample substrate.

II. SENSITIVITY ISSUES

Before focusing on the factors influencing the sensitivof the VSM, it is important to make some general remarThe literature presents several different numbers for thesitivity of various measurement systems without always csidering the differences in application of these systemsspecifying how the sensitivity number was obtained. Tsensitivity of a system should be presented as either theto peak or the root-mean-square~rms! noise floor~in units ofmagnetic moment! at a certain specified effective bandwidof the used electronics. However, the sensitivity is only revant if the actual useable maximal sample dimensionsknown, so that real magnetic signal to noise ratios foparticular magnetization can be compared. A system canpresented with a very high sensitivity, but if it is only suable for measuring very small samples~with very small sig-nals!, the measurement results obtained on this instrumare not necessarily better than those obtained with asensitive system on a much larger sample. If a systemmands the use of samples limited to, for example, 1 mm2 insize, it must be 100 times more sensitive than a systemwhich samples of 1 cm2 can be used, in order to producsimilar quality graphs. This is of course especially the cwhen the third dimension is limited such as in thin films.

If the optimal sample dimensions are known for a givconfiguration with a known performance, the dimensionsthe detection coil system can be scaled up or down protionally to fit other sample sizes, because almost all aspof a VSM system are scalable without changing the signto-noise ratio~SNR!.7

A review of several sensitivity issues has been givenRef. 7. The field noise problems as described by Richhave not been observed in the system that is treated hThis might be due to a more stable electromagnet posupply. If proper cabling, impedance matching, and ampcation is used, the system noise can be brought back clo

FIG. 1. Reversal process in ag-Fe2O3 audio tape at 80°. The arrows indicathe magnetization vector. The downward half of a hysteresis loop is shoThe horizontal direction represents the sample plane, the vertical direthe normal to the tape.


.

-ls

.n--r

eak

-reabe

ntsse-

in

e

fr-tsl-

nrre.er-to

~within 2–33! the theoretical limit presented by the electcal ~Johnson! noise. The Johnson noise is given bA4kTRD f , where R is the sensor impedance andD f theeffective noise bandwidth of the signal amplifier. Apart frothe electrical~random! noise, the vibrational noise, which igenerally a form of correlated noise, is a severe limitationthe sensitivity of a VSM system.

The detection coil system is extremely sensitive to flchanges in order to be able to measure the small flux chacaused by the sample motion. Therefore, vibrations ofdetection coils~which change the angle between the dettion coil axes and the magnetic field or move the coils inslightly inhomogeneous applied field! will also cause fluxchanges in the detection coils, depending on the fistrength. The use of coil pairs~connected in anti-series! willeliminate the signal produced by flux changes equapresent in both coils. Vibrations with the same frequencythe sample motion are the most bothersome as other freqcies are filtered out by the lock-in amplifier. These vibratiocan be caused by a mechanical coupling between the actu~a modified loudspeaker! and the detection coil system or bother sources of vibration around the VSM. The two mocommon ways to reduce this problem are the use of vibradampers between actuator and measurement system arigid coupling of the detection coils to the magnet. A disavantage of commonly used vibration dampers in or direcattached to the vibrator system, is that because of themass of the damped system and the low frequency ofvibrations ~,100 Hz!, the damping materials must be vecompliant. As a result the whole vibrator–sample holdersembly can more or less move freely, which gives rise tpoor definition of the exact sample position. In order to pvent this, the vibrator unit was mounted on a heavy tabwhich could be damped by much stiffer vibration dampeso that the sample positioning remained very constant.

A third way of vibration damping is the use of passiveactive anti-vibration elements. Some commercial systelike the Princeton Applied Research~PAR! and Oxford In-struments VSMs utilize passive spring mounted weigwhich, if properly tuned, will vibrate in anti-phase to thactuator and therewith can significantly reduce the induvibrations in the system. Here, rather than a passive aphase damper, an active system is used. This incorpora

n.on

FIG. 2. Schematic VSM setup including anti-phase vibrator and vibratdamping.


ithv

as

ectothhe

thCedr

ioan

caap

onithl,lan

itig

oerhetesase

-e

o

if

beth a

the

oleon-si-

n-rdt ofthea ofen-po-

the

byen-in a

n ormmnsi-

e-wedtiveils.iond is

plethe

r tor.gen ange-

.

3206 Rev. Sci. Instrum., Vol. 69, No. 9, September 1998 Samwel, Bolhuis, and Lodder

second actuator driving a dummy mass mounted on topthe first actuator and driven at the same frequency but wamplitude and phase tuned such that the vibrations insystem can be reduced by a factor of 1000 or better, ifbration feedback is used.

Both the vibration isolation elements and the anti-phvibration system are shown in Fig. 2.

III. VECTOR SIGNAL DETECTION

If a VSM is available with a detection coil system~DCS!suitable for the detection of the magnetization in one dirtion, the simplest thing to do in order to accomplish vecsignal detection, is to add a similar set perpendicular tofirst one. There are, however, several other approacwhich are discussed by Bernardset al. in Ref. 1. All standarddetection coil systems are generally connected stiffly topoles of the magnet in order to prevent vibrations of the Din the applied field. A change of the direction of the applifield with respect to the sample can be accomplished bytating the magnet–DCS combination around the sampleby rotating the sample inside the magnet–DCS combinatThe latter is generally easier due to weight of the magnetits electrical wiring and cooling hoses.

A sample that is relatively small compared to its distanto the detection coils can be considered as a dipole andproached as such. However, if the sample dimensionscomparable to or larger than the distance of the sam~edges! to the detection coils, the sensitivity of the detecticoil will depend on the exact position of the sample wrespect to the coils. If this~large! sample is nonsphericae.g., sheet shaped and rotated around one of the in-pdirections, then the geometric shape as seen by the sewill change with the angle and therefore the system sensity will be angle dependent. If the sample holder has an anwith the sensor~in the vertical direction!, rotation will causea circular motion of the sample with respect to the sensmaking the sensitivity even more angle dependent. Thfore, when the rotation is not exactly concentric within tcoils or when a relatively large sample is used, the sysneeds to be calibrated for each angle individually. This icumbersome and time consuming process that can elead to errors if, for example, a calibration sample is usthat cannot be saturated in all directions~due to the demagnetizing field!. A correct calibration procedure has been dscribed by Richter in Ref. 7 and Bolhuis in Ref. 8.

As the sensitivity of a DCS depends on the positionthe sample, the calibration for different angles will becomsensitive to errors in the positioning of the sample and

FIG. 3. Adapted Mallinson~see Ref. 10! setup compared to new coil setup


ofthe

i-

e

-res,

eS

o-orn.d

ep-rele

nesorv-le

r,e-

mailyd

-

fea

change in sample position occurs, the system needs torecalibrated. For this reason some VSMs are equipped wisample holder guidance.

Alternatively, one could choose to either increasedimensions of the coil system~in so far as this is possiblewithin the limitations presented by the electromagnet pgap! or decrease the size of the sample, which are functially equivalent. The disadvantage of doing this is that sentivity is sacrificed for obtaining a lower positional depedency. The sensitivity drops with approximately the thipower of the distance to the coil. Furthermore the amounsignal produced by a sample is directly proportional tomagnetic volume of the sample. This creates the dilemmmaking the system either very sensitive and angular depdent or making the system insensitive to sample shape,sition, and orientation, and therewith less sensitive tosignal as such.

Bernards6 has proposed a solution to this dilemmausing a 12 coil arrangement where the coils largely compsate each others positional dependency. This has resultedsystem that is more or less insensitive to the exact positiothe angular orientation of the sample for samples up to 6wide while at the same time preserving a reasonable setivity.

IV. A DIFFERENT DETECTION COIL SETUPAPPROACH

In order to prevent rather than minimize the abovdescribed problems, a different approach has been follohere. The described problems all arise due to the relarotation of the sample with respect to the detection coTherefore, in the new setup, the position of the detectcoils and sample are locked to each other and the fielrotated around both the sample and the sensor. Using simtrigonometry one can easily calculate the components ofmagnetization vector in the direction of and perpendiculathe applied field from the measured magnetization vecto

A similar arrangement was used in Ref. 9 to study imaeffects, but to our knowledge, it has never been used ipermanent setup as a magnetic thin film measuring arrament.

FIG. 4. Detection coil arrangement,Cb54, Cd514, CL53, Dx5Dy521.5Dz519 ~all in mm!.


o

ntivnce

m

vie

fo

ge

itdosilla

are

oto

nng

in

eenosi-ing

w,oilsh a

byt

plettrib-tion

the

th

a

mtwo

mpleolder.repro-llowlderlids.


If the magnet is to be rotated around the detection csystem, only a limited space for the detection coil systemavailable and the most obvious detection coil arrangemean orthogonal system. In order to create an equal sensitfor both sets of coils and in order to decrease the positiodependency of the sensitivity, the detection coils are plaunder a 45° angle with respect to the sample~see Fig. 3!. Thesignals in the sample plane and normal direction follow frosimple trigonometry.

A. Advantages and disadvantages of this coil setup

The advantages of the new setup follow from the preous section describing the problems in angle dependent msurements:

~i! There is only one calibration number necessaryeach set of coils;

~ii ! mispositioning of the sample only leads to the chanof the calibration factor, it has no influence on threlative sensitivity for different field angles.

One of the disadvantages of the system is its higher sensity to vibrational noise since the coils are not connectedrectly to the magnet. This problem has been solved by cstructional improvements as described before using pasvibration damping and active anti-phase vibration cancetion.

In standard detection coil systems where the coilsplaced close to the magnet pole faces, the image effecthances the sensitivity and is therefore not bothersome~aslong as the pole faces do not saturate completely!. In thesystem described here, the coils had to be placed close tsample and relatively far from the pole faces in orderreduce the image effect sufficiently~to approximately61%!.This is necessary since in this case, the rotating magwould otherwise cause an image effect that would chawith the angle.

B. Realization of the coil setup and its performance

Using the formulas given in Ref. 1 a DCS has beendeveloped based on a compromise between the followdesired specifications:

FIG. 5. Measured image effect in the experimental VSM. A and B aretwo orthogonal coil sets. When the magnet is at 45°6180° the A coils showa maximum image and the B coils a minimum, the B coils will showmaximum for magnet angles of245°6180°.


ilisis

ityald

-a-

r

e

iv-i-n-ve-

en-

the

ete

g

~i! low image effect~,2%!;~ii ! reasonable sensitivity;~iii ! low positional dependency.

The sensitivity increases with decreasing distance betwthe sample and the detection coil system, as does the ptional dependency. The image effect drops for decreassample–detection coil distance~the sensitivity towards theimage is approximately 1/r i

5/2, with r i the effective coil–image distance!. As only a limited, fixed pole shoe gap~50mm! was available, and the image effect was to be kept lothe compromise led to the system shown in Fig. 4. The cwere wound using 0.15 mm diam copper wire coated witthermo-setting epoxy.

The effect of the extra image signal can be monitoredrotating the magnet–pole shoes~at zero field, the magnecurrent compensating for the pole–shoe remanence! around asample in remanence. As the remanence of the samdoesn’t change, the change in measured signal can be auted to the image effect. The signal measured as a funcof the magnet angle for a~square! 1 cm2 sample in rema-nence is given in Fig. 5. The signal is normalized against

e

FIG. 6. Normalized measured positional sensitivity, obtained with a 1 c2

sample. The A and B direction are the directions of the coil axes of theorthogonal coil sets. The elipsoid subtends the region within 1% error.

FIG. 7. Sketch of the sample holder. The wedge shaped top of the saholder connector guarantees a reproducible placement of the sample hThe samples and substrates are placed in three slits guaranteeing aducible and stable positioning. If the sample holder itself has very shaslits and deeper slits are created in the ‘‘lid’’, than the same sample hocan be used for different sample thicknesses simply by using different


cial

3208 Rev. Sci. Instrum., Vol. 69, No. 9, September 1998 Samwel, Bolhuis, and Lodder

FIG. 8. Background curves for the new sample holder, the new sample holder when the anti-phase vibrator is switched off, and a standard commer~DMS!glass sample holder.

ex

fee

bth

bt

cmth

iiti

ure-theall

ither-

e

antt ofthin

dticeticos-areag-enples,kom

v-f

signal at 0°. Due to the 45° rotation of the system, thetrema in the graph are reached at 45°690°. One sees theexpected sensitivity behavior showing that the image efis indeed limited to only61%. The difference between thshape of the image effect curve and a sine function canattributed to the small differences between the coils andnot exactly centered sample.

The positional dependency of the sensitivity canmonitored by moving a remanent measurement sample inarea between the coils. The experimental results for a 12

thin film sample are shown in Fig. 6. One can see fromgray area in the figure that a positional error of60.5 mmwill cause a signal change of approximately 1.5%. Thisacceptable in this system however since the sample pos

FIG. 9. Background curves when using a single~clean! substrate or threesubstrates that largely compensate each other due to the design osample holder.


-

ct

ee

ehe

e

son

with respect to the coils does not change during the measment in contrast to other systems where the rotation ofsample around its axis is mostly accompanied by a smmotion of the rotation center. As the VSM is equipped wa micrometer precision positioning system, a positioningror of less than 0.5 mm can be accomplished.

The rms noise in the system is approximately 1 nA m2 ~1memu!, using a measurement time constant of 10 s~12 mHzeffective noise bandwidth!. This noise is independent of thapplied field, which indicates good vibration isolation.

At a maximum field~800 kA/m! the noise is virtually thesame, which indicates that field noise is not a significproblem in this system. Both noise figures are independensample size but are accomplished in a system that allowsfilm samples with a maximum size of 1 cm2.

V. DESIGN OF AN IMPROVED SAMPLE HOLDER FORTHIN FILM MEASUREMENTS

VSM measurements on thin films will often be diluteby the contribution to the signal of the moving diamagneor paramagnetic sample holder as well as by the magnproperties of the substrate on which the thin film is depited. For very low output samples substantial correctionsneeded in order to recover the signal coming from the mnetic layer under investigation. This is especially true whmeasurements are done on for example hard disk samwhere the~very thin! recording layer is deposited on a thicsubstrate of glass or aluminum. Background signals frhard disk substrates can be as high as 2mA m2 ~2 memu! ata field of 800 kA/m~;10 000 Oe!, which is in many caseslarger that the signal produced by the thin film itself. Reco

the


dds

yde

prl athiot

xem

bhes

slithrs

pliz

rs

yldF

intionuator

ag-fthethepty

r

canundof

lderxi-

t attee

n-


ering the measurement loop from such a signal can leaextra noise and problems if the measurement vector neebe retrieved.

As a separate issue, the reproducibility and accuracthe measurement depend to a high degree on the reproibility of the position and orientation of the sample in thsystem. We describe a sample holder which precisely reduces the position and orientation of the sample as welreduces the contribution of both holder and substrate tosignal measured. The reproducibility of the sample posithas been established by deepening the part in whichsample should reside. The sample is clamped to its fiposition by means of a lid and an elastic sleeve of the samaterial as the rest of the holder~Fig. 7!. This way the back-ground signal from the support rod has been minimizedkeeping its cross section as uniform as possible along tzaxis. The parts that are disturbing the uniformity of the crosection should be repeated symmetrically along thez axis ateach side of the centers of the coils. Therefore two moreare made, where the distance between the centers of theslits should be equal to the distance between the centethe coils along thez direction.

As a result there are two more positions in the samholder for putting in a sample. Substrates of the same sshape, and material as that on which the thin magnetic filmdeposited can be mounted in these two extra slits in ordeautomatically compensate for the substrate backgroundnal.

The material used for the sample holder is polymethmethacrylate~PMMA! which is fairly easy to machine anhas a smooth surface, which makes it easy to clean.cleaning the holder we use hydrochloric acid~30%! in anultrasonic cleaning bath for at least 1 h.


toto

ofuc-

o-se

nhede

y

s

tsreeof

ee,istoig-

-

or

The connector to the vibration unit has been madesuch a way that it exactly reproduces the angle of orientaof the holder. Because we use a loudspeaker based actthe distance between the actuator and the center of the mnet is quite large~.70 cm!. Therefore a carbon–fiber shawhich is very stiff, straight, and light, is used to extend tsample mount part towards the connector. As a resultbackground signal in a hysteresis curve, with the emsample holder installed, has a slope of less then 10 pA m2 perkA/m (7.958310210 emu/Oe). The background signal fothis new sample holder is roughly 103 lower than the back-ground signal for a commercial glass sample holder, asbe seen in Fig. 8. This figure also shows that the backgrosignal due to vibrations is significantly reduced by meansthe anti-phase vibrator. The empty substrates in the howill reduce the substrate background signal by appromately a factor of 10, as can be seen in Fig. 9.

ACKNOWLEDGMENTS

The research described in this article was carried outhe MESA Research Institute of the University of Twenwith financial support from Oxford Instruments, which thauthors gratefully acknowledge.

1J. P. C. Bernards, Rev. Sci. Instrum.64, 1918~1993!.2S. Foner, Rev. Sci. Instrum.27, 548 ~1956!.3S. Foner, Rev. Sci. Instrum.30, 548 ~1959!.4G. W. van Oosterhout, Appl. Sci. Res., Sect. B6, 101 ~1956!.5H. Plotkin ~unpublished 1951!, see S. Foner, Rev. Sci. Instrum.27, 548~1956!, Ref. 2.

6P. J. Flanders, M. M. M. Conference~1956!.7H. J. Richter, J. Magn. Magn. Mater.111, 201 ~1992!.8T. Bolhuis, L. Abelmann, J. C. Lodder, and E. O. Samwel, MRM Coference, J. Magn. Magn. Mater.~accepted!.

9S. R. Hoon and S. N. M. Willcock, J. Phys. E21, 480 ~1988!.10J. Mallinson, J. Appl. Phys.37, 2514~1966!.


Documents

An alternative approach to vector vibrating sample magnetometer